DETERMINATION OF FORMING LIMIT CURVES OF STEEL PIPES FOR HYDROFORMABILITY EVALUATION OF AUTOMOTIVE PARTS RAMIL KESVARAKUL A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR THE DEGREE OF MASTER OF ENGINEERING IN AUTOMOTIVE ENGINEERING (INTERNATIONAL PROGRAM) INTERNATIONAL COLLEGE KING MONGKUT’S INSTITUTE OF TECHNOLOGY LADKRABANG 2010 KMITL-2010-IC-M-004-007
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
DETERMINATION OF FORMING LIMIT CURVES OF STEEL PIPES FOR HYDROFORMABILITY EVALUATION OF AUTOMOTIVE PARTS
RAMIL KESVARAKUL
A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR THE DEGREE OF
MASTER OF ENGINEERING IN AUTOMOTIVE ENGINEERING (INTERNATIONAL PROGRAM)
INTERNATIONAL COLLEGE KING MONGKUT’S INSTITUTE OF TECHNOLOGY LADKRABANG
2010 KMITL-2010-IC-M-004-007
COPYRIGHT 2010
INTERNATIONAL COLLEGE
KING MONGKUT’S INSTITUTE OF TECHNOLOGY LADKRABANG
NATIONAL SCIENCE AND DEVELOPMENT AGENCY
Thesis Title: Determination of forming limit curves of steel pipes for
hydroformability evaluation of automotive parts
Student: Ramil Kesvarakul Student ID: 51061914
Degree: Master of Engineering
Programme: Automotive Engineering
Year: 2010
Thesis Advisor: Dr. Monsak Pimsarn Dr. Suwat Jirathearanat
Assoc.Prof. Naoto Ohtake
ABSTRACT
The aims of this research are to establish the forming limit curve (FLC) of tubular material
low carbon steels commonly used in Thai industry, verify these FLCs with real part forming
experiments, and compare these experimentally obtained FLCs against analytical ones available in
FEA software database. A self-designed bulge forming apparatus of fixed bulge length and a
hydraulic test machine with axial feeding are used to carry out the bulge tests. Loading paths resulting
to linear strain paths at the apex of the bulging tube are determined by FE simulations in conjunction
with a self-compiled subroutine. These loading paths are used to control the internal pressure and
axial feeding punch of the test machine. In this work a common low carbon steel tubing grade STKM
11A, with 28.6 mm outer diameter and 1.2 mm thick is studied. Circular grids are electro chemically
etched onto the surface of tube samples. Subsequently, the tube samples are bulge-formed. The
forming process is stopped when a burst is observed on the forming sample. After conducting the
bulge tests, major and minor strains of the grids located beside the bursting line on the tube surface
are measured to construct the forming limit curve (FLC) of the tubes. The forming limit curves
determined for these tubular materials are put to test in formability evaluations of test parts forming in
real experiment.
It was found that the tool geometry can keep the strain ratio constant is not dependent on the
thickness but only on OD of the tube, as in equations 𝐿 = 𝑂𝐷 and 𝑟 = ×. . The experimental
FLDs have predicted failures in forming process consistently with the real experiments. The
experimentally obtained forming limit curves (determined following STKM 11A) differ from
empirical one (from FEA software) and analytical one by about 0.02339 and 0.15736 true strain
respectively at FLD0, the corresponding plane strain values.
ACKNOWLEDGEMENT This thesis could not be completed without the assistance of many persons to whom I would
like to express my sincere appreciation.
First, I would like to sincerely thank my advisor, Dr. Suwat Jirathearanat, who has given me
many helpful suggestions, useful advice during the undertaken research.
I would also like to sincerely thank Asst. Prof. Dr.Monsak Pimsarn for kind advising and
helping, and Assoc. Prof. Naoto Ohtake for the suggestion of oxide analysis.
Moreover, I would like to show gratitude to the National Metal and Materials Technology
Center (MTEC) laboratory for providing the laboratory equipments and instruments as well as
financial supporting.
I am grateful to the National Science and Technology Development Agency (NSTDA), which
provided the full scholarship for studying in the master program.
Finally, I am very grateful to my family for all love, caring, understanding and motivation
throughout my life.
Ramil Kesvarakul
III
CONTENTS
Page
ABSTRACT I
ACKNOWLEDGMENTS II
CONTENTS III
LIST OF TABLES VII
LIST OF FIGURES VIII
CHAPTER 1 INTRODUCTION
1.1 Significance and Background 1
1.2 Objectives 2
1.3 Scopes 2
1.4 Expected Results 2
CHAPTER 2 THEORY AND LITERATURE REVIEWS
2.1 Introduction to Tube Hydroforming (THF) 3
2.2 Material Property 5
2.2.1 Stress 5
2.2.2 Strain 5
2.2.3 Tensile Test 6
2.2.4 The Engineering Stress-Strain Curve 6
2.2.5 Anisotropy 9
2.3 Tubular blank 10
2.4 Strain-based forming limit curve 11
2.4.1 Formulation of plastic instability criteria 14
2.4.2 FLC obtained by Finite Element software: DYNAFORM 20
2.5 Ductile fracture criterion 21
2.6 Effect of non-linearity of strain path 24
IV
CONTENTS (CONT.)
Page
CHAPTER 3 RESEARCH METHODOLOGY
3.1 Numerical Investigation 27
3.1.1 Test die insert design 27
3.1.2 Determination of loading paths by FE-simulations 31
3.2 Experimental Investigation 36
3.2.1 FLC testing apparatus 36
3.2.2 THF Test Specimens 37
3.2.3 Hydraulic press 38
3.2.4 Pressure system, Hydraulic cylinders and punches 39
3.3 Grid measurement 39
3.3.1 Digital Microscope 39
3.3.2 Grid Curvature 40
CHAPTER 4 EXPERIMENTATION AND RESULTS
4.1 Tube Hydraulic Bulge Test 43
4.1.1 Forming limit experiments with axial feeding 44
4.1.2 Forming limit experiments without axial feeding 49
4.1.3 Forming limit of welded seam 51
4.2 Grid Measurement 52
4.3 Construction of the Forming Limit Curve(FLC) 53
CHAPTER 5 COMPARISON AND VERIFICATION
5.1 Empirical FLC, Analytical FLC and Experimental FLC 59
5.2 Verification of Experimental FLC 62
5.2.1 Verification of Experimental FLC with actual bulge test
load path 62
5.2.2 Verification of Experimental FLC with a real automotive part 64
V
CONTENTS (CONT.)
Page
CHAPTER 6 CONCLUSION AND SUGGESTIONS
6.1 Conclusions 66
6.2 Suggestions for Future Work 67
REFERENCES 68
APPENDIX 70
Appendix A: International Conference: Determination of forming limit
curves of tubular materials for hydroformability
evaluation of automotive parts 70
Appendix B: Determination of forming limit curves of tubular 79
Appendix C: Tooling schematic 87
BIOGRAPHY 95
LIST OF TABLES
Table Page
3.1 Design of simulation matrix 28
3.2 A series of simulation 29
3.3 Simulation results 29
3.4 A series of simulation, L/OD=1 and rd/t=15 30
3.5 Result of Simulation, L/OD=1 and rd/t=15 31
4.1 Grid curvature 52
LIST OF FIGURES
Figures Page
2.1 Schematic illustration of the hydroforming of a bulge in a tube 3
2.2 Example tube hydroformed parts: a 2004 Ford F-150, chassis frame 4
VI
LIST OF FIGURES (CONT.)
Figures Page
2.3 Components of stress on element (Hosford and Caddell, 2007). 5
2.4 Typical tensile specimen (Marciniak et al., 2002). 6
2.5 Load-extension diagram for tensile test (Marciniak et al., 2002). 6
Where Z is equal to Zd and Zl, as given in Equations (2.15) and (2.16), for diffused necking and
localized necking, respectively. The critical minor principal strain can be obtained from ε2C=ξ ε1C . A flow chart for determining the forming limit strains is shown in Figure 2.15. At first, the
exponent of the yield function, m, the strain-hardening exponent, n, and the normal anisotropy, r,
of the material are input. After the strain ratio ξ is given, the stress ratio can be calculated by
is used. The critical major strains corresponding to different obtained critical strain pairs (ε2C, ε1C) for 1 > ξ >-0.5
18
Start
Input n, r, m values
Set strain ratio ξ=-0.5
ξ≥0
Calculate ZdUsing Eq.(6)
Calculate ZlUsing Eq.(7)
Calculate ε1c Using Eq.(16)
ε2c = ξ ε1c
ξ≥1
Plot FLC
Stop
ξ = ξ +0.1
Yes
No
Yes
No
Figure 2.15 Flow chart for determining the critical major and minor principal strains.
Figure 2.16(a) and (b) shows the forming limit curve and the yield locus, respectively,
using Hill’s non-quadratic yield function with r = 1.856 and n = 0.226. The region for stress ratios
(𝛼 = 𝜎 𝜎⁄ ) from 0.5 to 0 in the yield locus figure corresponds to that for strain ratios (𝜉 =𝜀 𝜀⁄ ) from 0 to -0.5 in FLC figure.
19
(a) Forming limit curve
(b) Yield locus
Figure 2.16 The forming limit curve and yield locus with Hill’s non-quadratic yield function.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5
𝜎 /𝜎
-1 1
1
-1
0.5
-0.5
0.5 -0.5
Min
or st
rain
( 𝜀)
Minor strain (𝜀 )
𝜎 /𝜎
20
2.4.2 FLC obtained by Finite Element software: DYNAFORM
FE simulations (Dynaform) are used to determine the loading paths in this study.
The analytical FLCs using the n and r values obtained by tensile tests. The FLC curve
approximately according to Keeler’s formula as given below:
𝐹𝐿𝐷 = ( . . ) , 0 < 𝑡 < 2.54 𝑚𝑚 ; (2.26)
𝐹𝐿𝐷 = [ ( . . ) ] , 2.54 ≤ 𝑡 ≤ 5.33 𝑚𝑚 ; (2.27)
The shape of FLC is determined by the formulas below:
𝜀 = 𝐹𝐿𝐷 + 𝜀 (0.027254𝜀 − 1.1965) 𝜀 < 0 (2.28)
𝜀 = 𝐹𝐿𝐷 + 𝜀 (−0.008565𝜀 − 0.784854) 𝜀 < 0 (2.29)
The critical major and minor strains are calculated Keeler’s formula and plotted
to construct the forming limit curve (FLC), as show in Figure 2.17.
Figure 2.17 The forming limit curve with Keeler’s formula.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5
Min
or st
rain
( 𝜀)
Minor strain (𝜀 )
21
2.5 Ductile fracture criterion
A metal that plastically deforms extensively without the onset of fracture is normally
termed as ductile. Ductile fracture occurs when a material is subjected to large deformation. In a
metal forming process, the ductile fracture is so complicated that many experimental
investigations as well as theoretical predictions have been performed to determine that of the
metal, and various criteria have been proposed to evaluate the forming limit (Cockroft M. 1968,
Norris DM. 1978, Brozzo P. 1972, McClintock FA. 1968, Oyane M. 1980). Unfortunately, it
appears that the various criteria have a number of validity restrictions and there are no universal
criteria for metal forming process. It is well known that the forming limit of hydroforming
process depends greatly upon the deformation history. Therefore, the histories of stress and strain
may have to be considered in the criteria. Among these criteria, the ductile fracture criterion
proposed by Oyane (1980) was successfully introduced in prediction of forming limit of a sheet
metal forming process by Takuda and others (Takuda H. 1999, Takuda H. 1999, Mori K. 1996,
Takuda H. 2000). The past researches show that Oyane’s ductile fracture criterion can be applied
to evaluate the forming limit in a wide range of metal forming processes, including aluminum
alloy material which does not have evident localized necking before fracture. Oyane’s ductile
fracture criterion assumes that the history of hydrostatic stress affects the occurrence of the
ductile fracture;
∫ + 𝐶 𝑑𝜀̅ = 𝐶 (2.30)
where 𝜀̅ is the equivalent strain at fracture, 𝜎 the mean stress, 𝜎 the equivalent stress, 𝜀 ̅the
equivalent strain, and 𝐶 , 𝐶 the material constants. In Equation (2.30), it is represented that the
fracture occurs when the value of the left-hand side reaches that of the right-hand side. Oyane’s
criterion requires two material constants 𝐶 and 𝐶 , which can be obtained from limit strains
corresponding to uniaxial tensile test and plane-strain tensile test.
Equation (2.30) can be rewritten as follows:
𝐼 = ∫ + 𝐶 𝑑𝜀 ̅ (2.31)
The histories of stress and strain in each element during forming are calculated by the FEM, and
22
the ductile fracture integral I in Equation (2.31) is obtained for each element. When the integral
value I of Equation (2.31) reaches 1.0, the fracture will occur. This ductile fracture value I can be
calculated for every finite element during the forming process.
Figure 2.18 Forming limit diagram obtained experimentally for a STKM-11A tube
(Li-Ping Lei, 2002).
Figure 2.18 shows the experimentally obtained forming limit diagram from bulge test of a
STKM-11A tube (Li-Ping Lei, 2002). The black dots indicate the strains just at the fracture site.
Namely, the ultimate strains for fracture are distributed in the figure. After the localized necking
occurs, the plastic deformation almost ceases outside of the necking, while the deformation at the
necking region progresses under plane-strain condition to fracture. It is found that the strains after
the localized necking tend to decrease with increase in the strain ratio, 𝛽 = 𝜀 /𝜀 , where 𝜀 and
𝜀 are the major and the minor strains on the tube. As a result, the solid circles are linearly
distributed as shown in Figure 2.18.
According to the Mises yield criterion and the Levy–Mises stress–strain increment
relationship, the terms in Equation (2.30) are expressed by the strain ratio, 𝛽, as
= ( ) = 𝐴(𝛽), (2.32)
𝑑𝜀̅ = (1 + 𝛽 + 𝛽 )𝑑𝜀 = 𝐵(𝛽)𝑑𝜀 (2.33)
Min
or st
rain
( 𝜀)
Minor strain (𝜀 )
23
Figure 2.19 depicts the above relations. The rations 𝜎 𝜎⁄ and 𝑑𝜀̅ 𝑑𝜀⁄ increase with 𝛽.
Since the unknowns are two material constants 𝐶 and𝐶 , two arbitrary fracture
strains𝜀 (𝛽 ),𝜀 (𝛽 ) in Figure 2.18 should be chosen to get these values. Provided that the
strain ratios (𝛽) are constant during the deformation until the fracture initiation, the material
constants are simply calculated from Equation (30), (32) and (33) as the following:
𝐵(𝛽 )𝜀 (𝛽 ) −1𝐵(𝛽 )𝜀 (𝛽 ) −1
𝐶𝐶 = −𝐴(𝛽 )𝐵(𝛽 )𝜀 (𝛽 )
−𝐴(𝛽 )𝐵(𝛽 )𝜀 (𝛽 ) (2.34)
Figure 2.19 Variations of 𝜎 𝜎⁄ and 𝑑𝜀̅ 𝑑𝜀⁄ with respect to strain ratio 𝛽 (Li-Ping Lei, 2002).
Figure 2.20 Determination of C1 and C2 for a STKM-11A tube (Li-Ping Lei, 2002).
Min
or st
rain
( 𝜀)
Minor strain (𝜀 )
24
The material constants C1 and C2 are determined approximately so that the fracture
strains calculated for constant ratios 2t the experimental ones as appeared in Figure 2.20 and then
C1 =-0.069, C2 =0.255 for STKM-11A material are obtained. The calculated fracture strains are
distributed linearly for each pair of material constants. Hence it is found that the distributions of
the fracture strains calculated from the ductile fracture criterion are similar to those of the black
dots as shown in Figure 2.20.
2.6 Effect of non-linearity of strain path.
Several researchers have shown experimentally that a non-linear strain path can change
the shape and location of the FLC in strain space. Ghosh and Laukonis (1976) investigated sheet
metal forming limit curve based on plastic deformation energy by prestraining stainless steel
specimens to various levels of strain and in different deformation. Graf and Hosford (1993a,
1993b) also reported strain-path effects for 2008 T4 aluminium pre-strained in uniaxial, equi-
biaxial and near plane-strain tension. The result was a different FLD for each value of prestrain as
shown in Figure 2.21. Graf and Hosford (1993a) also attempted to predict the shifted FLD using
the Marciniak-Kuczynski (M-K) analysis; however, the predicted FLC did not correlate well with
experimental data when the prestrain was in equi-biaxial tension.
Figure 2.21 Influence of strain path on the FLC (adapted from Graf and Hosford, 1993).
25
These experimental observations show that, depending on the loading history, the actual
FLC can be significantly different from the as-received FLC. As a result of a change in shape and
position, combinations of principal strains that are safe from necking can lie above the as-
received FLC, and conversely, failures can take place at strains below the as-received FLC.
Furthermore, during any forming operation, different locations on a part undergo different strains
and forming modes. If the component is manufactured in two or more forming stages, the overall
strain path in each location can be severely non-linear as a result of following one strain path in
one forming stage and a different strain path in the next forming stage. In order to obtain a
standardized assessment of the forming limit curve, the strain path of the bulged sample has to be
controlled as linear as possible (i.e., constant strain ratio). This would call for one and the same
material to evaluate the forming severity of parts that were produced in complex, multi-stage
forming processes such as are typical for hydroformed tubular components.
CHAPTER 3
RESEARCH METHODOLOGY
This research aimed to establish the forming limit curve (FLC) of tubular material low
carbon steels commonly used in Thai industry, namely STKM 11A with 1.2 mm thickness and
28.6 mm outer diameter, compare these experimentally obtained FLCs against analytical and
empirical ones available in FEA software (Dynaform) and Formulation of plastic instability
criteria, and verify these FLCs with real part hydroforming experiments. First, bulge forming
apparatus of fixed bulge length and a hydraulic test machine with axial feeding were used to carry
out the bulge tests that were able to keep linear strain part at the apex of the bulging tube. FEM
was used to determine die proper insert parameters such as free bulge length (L), die entry radius
(rd). Second, loading paths corresponding to the strain paths with constant strain ratios at the apex
of the bulging tube were also determined by FE simulations, which in turn were used to control
the internal pressure and axial feeding punch of the test machine. Third, preparing the tube blank,
circular grids were electro chemically etched onto the surface of tube samples. Then, after
running bulge tests, the major and minor strains of the grids beside the bursting line on the tube
surface are measured to construct the forming limit curve. Finally, compare these FLCs against
empirical FLCs and verify these FLCs with real part forming experiments to achieve the
objectives and scope of this research. The procedure of research is as follows.
1. Study tools and equipment used in research.
2. Study material properties used in the experiments.
3. Study the forming process.
4. Learn how to use Dynaform.
5. Run Tube Hydroforming simulations for test die insert design and process parameter
determination.
6. Prepare Tube blank such as cutting, sanding, griding of any sharp edges and electro
chemically etched onto the surface of tube samples.
7. Conduct the experiment, collect and analyze the data.
8. Construct the forming limit curve.
9. Compare and verify the obtained FLD’s.
27
3.1 Numerical Investigation
3.1.1 Test die insert design
Several researchers (Jieshi Chen, Xianbin Zhou and Jun Chen, 2009) have
shown experimentally that non-linear strain path can change the shape and location of the FLC.
Nevertheless, to an extent, this effect on the FLC can be minor if the non - linearity is kept small.
In order to obtain a standardized assessment of the forming limit curve, the strain path at the apex
of the bulged sample has to be controlled as linear as possible (i.e., constant strain ratio). In this
work, it was necessary to design proper testing die insert geometry – 1)die entry radius(rd) and 2)
bulge length(L), as show in Figure 3.1.
Figure 3.1 Schematic diagram of the test die insert parameters
An FEA software (DYNAFORM) was used to conduct all the simulations in this
work. Due to its symmetry, only one half of the testing a die insert and tube sample were model,
see Figure 3.2. In each case of simulation run, several process parameters (i.e. pressure and axial
Free bulge length (L)
Initial tube blank
Outer diameter (OD)
Initial tube thickness (t)
Axial Punch
Die entry radius (rd)
Axial feed Axial feed
Pressurized water inlet
Pressure
28
feed distance) were tried in an attempt to form the sample with linear strain paths. Any elements
in the red indicate that the parts have failed.
Figure 3.2 The formed part show in half model
The specimen tubes were modeled with 1 mm thickness and 25.4 mm outer
diameter. Two geometry parameters were used to investigate, namely 1) L/OD and 2) rd/t, see
table 3.1. A series of simulation were conducted with various die insert geometry and tube
sample dimensions, see table 3.2, to determine the proper testing die insert geometer.
Table 3.1 Design of simulation matrix
L/OD 1 2 3
L 25.4 50.8 76.2
OD 25.4 25.4 25.4
rd/t 5 15 25
rd 5 15 25
t 1 1 1
𝑟
𝑡 2
29
Table 3.2 A series of simulation
Model L/OD rd/t L OD rd t
RUN01 1 5 25.4 25.4 5 1
RUN02 1 15 25.4 25.4 15 1
RUN03 1 25 25.4 25.4 25 1
RUN04 2 5 50.8 25.4 5 1
RUN05 2 15 50.8 25.4 15 1
RUN06 2 25 50.8 25.4 25 1
RUN07 3 5 76.2 25.4 5 1
RUN08 3 15 76.2 25.4 15 1
RUN09 3 25 76.2 25.4 25 1
It was found that only properly designed die insert geometry relative to tube
sample geometry – 1) tube outer diameter (OD) and 2) tube sample thickness (t) will allow the
linear strain path during testing. Four strain ratios (𝜉 = 𝜀 𝜀⁄ ) -0.1, -0.2, -0.3 and -0.4 were the
slopes of each strain path under investigation in this work. It no possible internal pressure and
axial feed distance were found to fulfill the constant strain ratios then it was concluded that the
specific die insert geometry is not proper.
Table 3.3 Simulation results
Model L/OD rd/t L(mm) OD(mm) rd(mm) t(mm) Resultant strain ratio(𝜉)
RUN01 1 5 25.4 25.4 5 1 Non-linear
RUN02 1 15 25.4 25.4 15 1 Linear
RUN03 1 25 25.4 25.4 25 1 Non-linear
RUN04 2 5 50.8 25.4 5 1 Non-linear
RUN05 2 15 50.8 25.4 15 1 Non-linear
RUN06 2 25 50.8 25.4 25 1 Non-linear
RUN07 3 5 76.2 25.4 5 1 Non-linear
RUN08 3 15 76.2 25.4 15 1 Non-linear
RUN09 3 25 76.2 25.4 25 1 Non-linear
30
Table 3.3 summarizes the results of this simulation finding. Only die insert
geometry in simulation RUN02 yields the satisfying linear strain path requirement (i.e. constant
strain ratio𝜉 ± 0.01). It can be seen that only RUN02 is able to keep the strain ratio of -0.4
constant (i.e. linear) all the way to the fracture limit, FLC.
The ratio of Model Run02 (L/OD=1 and rd/t=15), were put to test in another
outer diameter and thickness of tube sample. A series of simulation were conducted with two
geometry ratio to verify, namely 1) L/OD=1 and 2) rd/t=15, see table 3.4, to verify the proper
testing die insert geometer.
Table 3.4 A series of simulation, L/OD=1 and rd/t=15
Model L/OD=1 rd/t=15
L(mm) OD(mm) rd(mm) t(mm)
TEST01 12.7 12.7 7.5 0.5
TEST02 12.7 12.7 15 1
TEST03 12.7 12.7 30 2
TEST04 25.4 25.4 7.5 0.5
TEST05 25.4 25.4 15 1
TEST06 25.4 25.4 30 2
TEST07 50.8 50.8 7.5 0.5
TEST08 50.8 50.8 15 1
TEST09 50.8 50.8 30 2
Table 3.5 summarizes the results of second simulation finding. Die insert
geometry in simulation TEST01, TEST05 and TEST09 yields the satisfying linear strain path
requirement (i.e. constant strain ratio𝜉 ± 0.01). It was found the tool geometry that can keep the
strain ratio constant is not dependent on the thickness but dependent on only OD of the tube, as
given in Equations (3.1)-(3.2), see figure 3.3.
31
Table 3.5 Result of Simulation, L/OD=1 and rd/t=15
Model L/OD=1 rd/t=15 Results
L(mm) OD(mm) rd(mm) t(mm)
TEST01 12.7 12.7 7.5 0.5 Linear
TEST02 12.7 12.7 15 1 Non-linear
TEST03 12.7 12.7 30 2 Non-linear
TEST04 25.4 25.4 7.5 0.5 Non-linear
TEST05 25.4 25.4 15 1 Linear
TEST06 25.4 25.4 30 2 Non-linear
TEST07 50.8 50.8 7.5 0.5 Non-linear
TEST08 50.8 50.8 15 1 Non-linear
TEST09 50.8 50.8 30 2 Linear
𝐿 = 𝑂𝐷 (3.1)
𝑟 = ×. (3.2)
Figure 3.3 testing die insert geometry
3.1.2 Determination of loading paths by FE-simulations
The ratio of Proper Model according to equations (3.1) and (3.2), were put to
model with 1.2 mm thickness and 28.6 mm outer diameter tube blank. Internal pressure and axial
feed distance under investigation in this work were used to control the four strain ratios (𝜉 =𝜀 𝜀⁄ ) -0.1, -0.2, -0.3 and -0.4 at the apex of the bulged sample as linear as possible. The
corresponding loading paths to the four linear strain paths are determined by FE simulations,
shown in Figure 3.4, 3.5 and 3.6. Results of the formed simulation with four different strain ratios
Outer diameter (OD)
𝑟 = 15 × 𝑂𝐷25.4
𝐿 = 𝑂𝐷
32
(𝜉 = 𝜀 𝜀⁄ ) -0.1, -0.2, -0.3, -0.4 and no feeding show in half model, see Figure 3.7 and each of
Figure 3.6 Feeding distance (mm) with Internal Pressure (bar)
Figure 3.7 a.) No feeding
Figure 3.7 b.) 𝜉(𝜀 𝜀 ) = −0.1⁄
0
50
100
150
200
250
300
0 2 4 6 8 10
𝛏=-0.1 𝛏=-0.2 𝛏=-0.3 𝛏=-0.4 In
tern
al P
ress
ure
(bar
)
Feeding distance (mm)
35
Figure 3.7 c.) 𝜉(𝜀 𝜀 ) = −0.2⁄
Figure 3.7 d.) 𝜉(𝜀 𝜀 ) = −0.3⁄
Figure 3.7 e.) 𝜉(𝜀 𝜀 ) = −0.4⁄
Figure 3.7 Simulation results with four strain ratios (𝜉 = 𝜀 𝜀⁄ ) -0.1, -0.2, -0.3, -0.4 and no
feeding
36
Figure 3.8 Different strain paths investigated
3.2 Experimental Investigation
3.2.1 FLC testing apparatus
A bulge test apparatus with a fixed bulge length without axial feeding as shown
in Figure 3.9 is used to implement the forming limit experiments to obtain the strain path on right
side of the FLD. The specimen 300 mm initial tube lengths are expanded at both the end of the
tube and fixed by punches during forming.
Figure 3.9 The experimental apparatus for bulge tests without axial feeding.
ξ=-‐0.1
Maj
or st
rain
( 𝜀)
Minor strain (𝜀 )
ξ=-‐0.2 ξ=-‐0.3
ξ=-‐0.4
No feeding No feeding No feeding
1.00
0.80
0.60
0.40
0.20
0.00 -0.1 -0.5 -0.3 -0.1 -0.3 -0.5
37
A hydroforming test machine with axial feeding is used to conduct the
experiments with 200 mm initial tube length to obtain the strain paths on the left side of
the FLD, in which tensile and compressive strains occur as shown in Figure 3.10
Figure 3.10 The experimental apparatus for bulge tests with axial feeding.
Figure 3.11 Schematic diagram of the experimental apparatus for bulge tests
3.2.2 THF Test Specimens
Before bulge tests, the tubes of STKM low carbon steels are 200 mm (with axial
feeding) and 300 mm (without axial feeding) long, 1.2 mm thick, and 25.8 mm in outer diameter.
They are rounded off of any sharp edges in the tubular blanks by lathe machine, for the tubes used
for the forming limit experiments, circular grids with a diameter of 2.5mm as shown in Figure
3.12 are electrochemically etched on the tube surface before the experiments, see Figure 3.12.
38
Figure 3.11 Circular grids with a diameter of 2.5mm
Figure 3.12 THF Test Specimens
3.2.3 Hydraulic Press
In the hydroforming process, hydraulic presses are typically used to open and
close the die and to provide enough clamping load during the forming period to prevent die
separation. A 200 ton hydraulic press was used in this experiment. The press is controlled by a
CNC controller shown in Figure 3.13.
Figure 3.13 Hydraulic press and CNC controller
39
3.2.4 Pressure system, Hydraulic cylinders and punches
The pressure system (pump, intensifier and control and relief valves, coolers,
etc.) provide the required pressure levels, which are controlled by CNC controller, shown in
figure 3.14. The axial punches are necessary to seal the end of the tube to avoid pressure losses
and to feed material into expansion regions. They should feed the material in a controlled path,
and in synchronization with internal pressure.
Figure 3.14 CNC controller used to control internal pressure and axial punches
3.3 Grid measurement
After bulge tests, Dimensions of the grid circles at the pole (-45˚ to 45˚from welding
line) were accurately measured to obtain true major and minor strains by Equations. (2.8)- (2.9).
The critical major and minor strains are plotted to construct the forming limit curve (FLC) for
tubular material.
3.3.1 Digital Microscope
A Digital Microscope that shown in Figure 3.15.is used to take snapshots of the
deformed grids, which are later measured for true major and minor strains. After the bulge tests,
the measured major and minor strains on the tube surface are used to construct the forming limit
curve. A grid is captured by Dino Capture Software and measure the diameter of deformed grids
as show in Figure 3.16
Figure 3.15 Digital Microscope (Dino-Lite)
40
Figure 3.16 The deformed grids measured using Dino Capture Software
3.3.2 Grid Curvature
As specimen are tubular and final from are bulge shape, the captured grid from
Digital Microscope are projected from real deformed grids; therefore, corrective curve is
important to achieve the real deformed grid diameter and FLC. The curvature has two radii, 𝑟 is
meridian radius of curvature at the pole as shown in figure 3.17 and 𝑟 is circumferential radius
of curvature at the pole as shown in figure 3.18.meridian and circumferential radii of curvature at
the pole
Figure 3.17 𝑟 is meridian radius of curvature at the pole
Figure 3.18 𝑟 is circumferential radius of curvature at the pole
𝑙 ′
𝑟
𝑟
𝑙 ′
41
The digital camera with macro lens is used to capture the shape of bulge tube,
see in figure 3.19 and approximate the curve by CAD software, see in figure 3.20.
Figure 3.19 A photo of bulged tube for curvature measurement
Figure 3.20 Approximation of the curve by CAD software
Then, one can calculate the real deformed grid diameter by following equations.
𝑙 = 2𝑟 sin (3.1)
𝑙 = 2𝑟 sin (3.2)
Where 𝑙 = Major deformation
𝑙 = Minor deformation
𝑙 ′ = Captured major deformation
𝑙 ′ = Captured minor deformation
𝑟
𝑟
CHAPTER 4
EXPERIMENTATION AND RESULTS
This chapter explains experimentation and results of the experiments. A hydroforming test machine with axial feeding is designed for the left-hand side data in the forming limit diagram, in which tensile and compressive strains occur. Figure 4.1 shows a hydroforming test machine with axial feeding tooling set. A bulge test apparatus with a fixed bulge length without axial feeding as shown in Fig. 4.2 is used to implement the forming limit experiments to obtain the strain path on right side of the forming limit diagram. This test machine consists of three main parts for supporting the tooling; a hydraulic power system for providing the pressure source of the internal pressure and the feeding punches; and a control system.
Figure 4.1 Apparatus with feeding tooling set
Figure 4.2 Apparatus without feeding tooling set
43
4.1. Tube Hydraulic Bulge Test
A control system is used to control the forming pressure and the left and right axial feeding distances of the test machine according to the loading paths shown in Figure 4.3-4.4 and tryout in actual process.
Figure 4.3 Loading path, Feeding distance (mm) and Internal pressure (bar) with time(s)
4.1.1. Forming limit experiments with axial feeding
The actual responses of the forming pressures and axial feeding distances during the forming limit experiments are observed and recorded by computer PC-based data logger. Then, grouped correspond actual loading path specimen and separated off the non-corresponding actual loading path specimen. The actual responses are showed in Figure 4.6 and the results of the products after bulge tests with axial feeding for different strain ratios are shown in Figure 4.7. It is known that cracks or bursting lines occur around the pole of the bulged tubes and the maximum bulge height increases with the increase of the absolute value of the strain ratio
Figure 4.6 Results of the formed product for different strain paths.
0
50
100
150
200
250
300
350
0 2 4 6 8 10 12
Actual loading path-Set1Actual loading path-Set2Actual loading path-Set3Actual loading path-Set4
Intern
al Pre
ssure
(bar)
Feeding distance (mm)
Set1 Set2 Set4 Set3
Set1 (1.66, 330.91) Set2 (3.33, 323.68)
Set3 (7.25, 308.04)
Set4 (9.57, 262.63)
45
In order to consider the strain path, several forming runs were conducted with the same loading path but were stopped at different deformed states. The set 1-4 loading paths of different deformed state are showed in Figure 4.7-4.10 respectively. First experiment, Feeding distance 1.66 mm is applied at the ends of tube and actual pressure up to 330 bars which is more than expandability of the welding seam; thus the tube burst at welding seam before localized necking occurred. Due to small feeding distance, it was hard to control the end feeding distance and corresponding pressure as results shown in Figure 4.7.
a.) loading path of the formed product of different deformed state-Set1
b.) Results of the formed product of different deformed state-Set1
Figure 4.7 loading path and Results of the formed product of different deformed state-Set1
0
50
100
150
200
250
300
350
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8
Set1-01Set1-02Set1-03Set1-04Set1-05Int
ernal
Pressu
re (ba
r)
Feeding distance (mm)
Set1-01 (1.66, 330.91) Set1-02
(1.51, 308.65)
Set1-03 (1.23, 313.61)
Set1-04 (1.21, 271.77)
Set1-05 (0.38, 237.44)
Set1-01 Set1-02 Set1-04 Set1-03 Set1-05
46
a.) loading path of the formed product of different deformed state-Set2
b.) Results of the formed product of different deformed state-Set2
Figure 4.8 loading path and Results of the formed product of different deformed state-Set2
0
50
100
150
200
250
300
350
0 0.5 1 1.5 2 2.5 3 3.5 4
Set2-01Set2-02Set2-03Set2-04Set2-05Set2-06
Intern
al Pre
ssure
(bar)
Feeding distance (mm)
Set2-01 Set2-02 Set2-04 Set2-03 Set2-05
Set2-01 (3.33, 323.68)
Set2-03 (2.52, 323.67)
Set2-02 (2.74, 328.21)
Set2-04 (2.48, 327.48)
Set2-05 (2.28, 325.12)
Set2-06 (2.05, 323.91)
Set2-06
47
a.) loading path of the formed product of different deformed state-Set3
b.) Results of the formed product of different deformed state-Set3
Figure 4.9 loading path and results of the formed product of different deformed state-Set3
4.1.2. Forming limit experiments without axial feeding
A bulge test apparatus with a fixed bulge length without axial feeding are installed for running experiment. The actual responses are showed in Figure 4.11 and corresponding specimen are showed in Figure 4.12 .
Figure 4.11 Loading path, Internal Pressure (bar) with times(s)
Figure 4.12 Results of the formed product without axial feeding. In order to consider the strain path, several forming runs were conducted with the same loading path but were stopped at different deformed states. The set 5 loading path of different deformed state are showed in Figure 4.13.
0
50
100
150
200
250
300
350
0 1 2 3 4 5 6 7 8 9 10
Actual loading path-Set5
Prescribed loading path-set5
Intern
al Pre
ssure
(bar)
Times(s)
50
a.) loading path of the formed product of different deformed state-Set5
b.) Results of the formed product of different deformed state-Set5
Figure 4.13 loading path and Results of the formed product of different deformed state-Set5
0
50
100
150
200
250
300
350
400
0 2 4 6 8 10 12
Set5-01Set5-02Set5-03Set5-04Set5-05Int
ernal
Pressu
re (ba
r)
Times(s)
Set5-01 Set5-02 Set5-04 Set5-03 Set5-05
Set5-01 (335.14)
Set5-03 (332.01)
Set5-02 (332.40) )
Set5-04 (318.01)
Set5-05 (311.35)
51
4.1.3. Forming limit of welded seam
During ERW tube production, the two edges are welded together by electrical resistance welding (ERW). Because of low quality control of electrical resistance welding process, some specimens burst at welded seam as show in figure 4.14.
a.) loading path of the formed product of specimens that burst at welded seam
b.) Results of the formed product of specimens that burst at welded seam
Figure 4.14 The specimens that burst at welded seam and corresponding load path.
0
50
100
150
200
250
300
350
0 1 2 3 4 5 6 7 8
Set1Set2Set3Set4
Set1 Set2 Set4 Set3
Intern
al Pre
ssure
(bar)
Feeding distance (mm)
52
4.2. Grid Measurement
In order to measure the real deformed length, the grid curvatures are investigated. The digital camera with macro lens is used to capture the shape of bulge tube, which is later used to approximate the curve by CAD software as the result is showed in Table 4.1.
The deformed grid is measured ± 45 degrees around the welding line as shown in Figure 4.15.
Figure 4.15 Measure zone covering ±45 degrees from welding seam
4.3. Construction of the Forming Limit Curve(FLC)
The critical major and minor strains are plotted to construct the forming limit curve (FLC). Figure 4.16 show forming limit curve and the major and minor strains of all specimens covering ±45 degrees from the weld line.
Figure 4.16 FLC with the major and minor strains of all specimens covering ±45 degrees from welding seam
The deformed characteristic of specimens from welding line to 35 degrees is showed in Figure 4.17. The major and minor strains are smallest at 0-7.5 degrees from welding line, then increasing at 7.5-20 degrees and maintain at 20-35 degree. As a result, the maximum strain between 20 and 35 degree is the representation of strain between 35 and 90 degrees.
a.) The Set4 major strain with degree from welding line
b.) The Set4 minor strain with degree from welding line
Figure 4.17 The major and minor strain with degree from welding line of Set4
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0 5 10 15 20 25 30 35
Set4-01
Set4-02
Set4-03
Set4-04
Set4-05
Set4-06
Set4-07
Set4-08
Set4-09
-0.3
-0.25
-0.2
-0.15
-0.1
-0.05
00 5 10 15 20 25 30 35
Set4-01
Set4-02
Set4-03
Set4-04
Set4-05
Set4-06
Set4-07
Set4-08
Set4-09
Major
strain
(ε1)
Mino
r stra
in (ε
2)
Degree from welding line (Degree)
Degree from welding line (Degree)
Bursting Point
Bursting Point
55
Several researchers (Jieshi Chen, Xianbin Zhou and Jun Chen, 2009) have shown experimentally that non-linear strain path can change the shape and location of the FLC. Nevertheless, to an extent, this effect on the FLC can be minor of the non linearity is kept small. In order to obtain a standardized assessment of the forming limit curve, the strain path at the apex of the bulged sample has to be control as linear as possible(i.e., constant strain ratio). The strain paths at the pole of the forming tube for different strain ratios are shown in the figures 4.18-4.22.
Figure 4.18 Set1 strain path at 30 degree
Apparently, the least feeding distance used to express the excessive pressure which more than expandability of welding line, consequently the burst appeared at the welding line before a material localized necking appeared; therefore, the critical strain at above condition could not reach to real critical strain.
-0.029821, 0.332867
-0.022583, 0.294567
-0.016905, 0.168228
-0.009568, 0.117859 -0.006455, 0.100914
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
-0.035 -0.03 -0.025 -0.02 -0.015 -0.01 -0.005 0
Major
strai
n(ε 1
)
Minor strain (ε2)
56
Figure 4.19 Set2 strain path at burst degree (20 Degree) from welding line
Figure 4.20 Set3 strain path at burst degree (22.5 Degree) from welding line
Figure 4.21 Set4 strain path at burst degree (27.5 Degree) from welding line
Figure 4.22 Set5 strain path at burst degree (25 Degree) from welding line
-0.238818, 0.659100
-0.157375, 0.450786
-0.124211, 0.284021 -0.119721, 0.314686
-0.114675, 0.315154 -0.103640, 0.316034
-0.097549, 0.236801 -0.075012, 0.203583
-0.048021, 0.159593
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
-0.3 -0.25 -0.2 -0.15 -0.1 -0.05 0
0.051891347, 0.356595627
0.046204714, 0.352134881
0.039244886, 0.345964341
0.031308471, 0.203006774
0.009871323, 0.132266279
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0 0.01 0.02 0.03 0.04 0.05 0.06
Major
strai
n (ε 1
)
Minor strain (ε2)
Minor strain (ε2)
Major
strai
n (ε 1
)
58
From the experimental data, the FLD and forming limit curves (FLC) of STKM 11A tubes are constructed as shown in Figure 4.23. The strain paths at the pole of the forming tube for different strain ratios are also shown in the figure. The four strain paths are obtained from experimental data. As the approximate trend lines, the strain paths at the pole of the forming tube for different strain ratios have a trend line of data as linear. The critical strain from the experimental data and the analytical results, the FLD and forming limit curves (FLC) of STKM 11A tubes are constructed as shown in Figure 4.23.
Figure 4.23 The Experimental FLD and forming limit curves (FLC) of STKM 11A tubes
The weld seam in a tube leads to an obvious non-homogeneity in material properties, and the formability of the weld seam and its heat-affected zone is usually lower than that of the tube, so that tube failures occur closely or on the weld line as shown in Figure 4.23.
-0.081826, 0.396554
-0.153728, 0.516244
-0.238818, 0.659100
-0.029821, 0.332867 0.051891, 0.356596
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
-0.3 -0.25 -0.2 -0.15 -0.1 -0.05 0 0.05 0.1
Major
strai
n (ε 1
)
Minor strain (ε2)
At 27.5˚ from the weld seam
At 22.5˚ from the weld seam
At 20˚ from the weld seam
At 30˚ from the weld seam
At 25˚ from the weld seam
Forming Limit Diagram
Forming Limit Diagram of low quality weld seam tube
CHAPTER 5
COMPARISON AND VERIFICATION
5.1 Empirical FLC, Analytical FLC and Experimental FLC
The formulation of plastic instability criteria and Keeler’s formula are analytical FLC
and empirical FLC which are used to compare with the experimental FLC, see Figure 5.1. The
formulations used are given in chapter 2. The n value of the flow stress obtained from tensile tests
is used to construct the analytical and empirical FLC of STKM 11A tubes. From the chart, the
experimental FLC is quite close to the Keeler’s formula. The Formulation of plastic instability
criteria gives the lowest FLC.
Figure 5.1 Comparison of predicted forming limit strains with the experimental forming limit
strains for low carbon steels STKM 11A.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
-0.3 -0.2 -0.1 0 0.1 0.2 0.3
Keeler’s formula
Formulation of plastic instability criteria
Experimental FLC
Major
strai
n (ε 1
)
Minor strain (ε2)
60
Figure 5.2 Effects of the r value on the forming limit curve with Hill’s non-quadratic yield
function.
Figure 5.3 Effects of the n value on the forming limit curve with Hill’s non-quadratic yield
function.
Analytical and empirical FLC can be influenced significantly by material properties used.
Figure 5.2 shows the effect of the normal anisotropy of the material, r, on the forming limit curve,
using Hill’s non-quadratic yield function with m= 2.0 and n = 0.2. A larger r value can give a
small raise to the forming limit curve in the tensile–tensile strain region as shown in Figure 5.2. In
the tensile–compressive strain region, however, the forming limit curves are not influenced by the
-0.1
0.0
0.1
0.2
0.3
0.4
0.5
0.6
-0.2 -0.2 -0.1 -0.1 0.0 0.1 0.1
r=1.00, n=0.2
r=1.75, n=0.2
r=2.50, n=0.2
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
-0.2 -0.2 -0.1 -0.1 0.0 0.1 0.1
r=1.75, n=0.2
r=1.75, n=0.3
r=1.75, n=0.4
61
r value. It seems that the forming limit curves in the tensile–compressive strain region using
Hill’s localized necking criterion are not influenced by the m and r values in the Hill’s non-
quadratic yield function.
Figure 5.3 shows the effects of the strain-hardening exponent of the tube material, n, on
the forming limit curves using Hill’s nonquadratic yield function with m= 2 and r = 1.75. It is
apparent that the forming limit curves are influenced significantly by the n value. A material with
a larger n value undergoes larger plastic deformation before necking occurs, accordingly a larger
n value raises the forming limit curves. Based on its formulation, however, it is apparent that the
forming limit curve by the formulation of plastic instability criteria is not dependent on the
material thickness.
As a result of Comparison of predicted forming limit strains with the experimental
forming limit strains for low carbon steels STKM 11A, the Keeler’s FLC lies above the
formulation of plastic instability criteria and lies below the experimental FLC. Swift’s diffused
necking criterion and Hill’s localized necking criterion associated with Hill’s non-quadratic yield
function are developed to derive the critical principal strains at the onset of plastic instability,
while the predicted strains at the onset of necking are always smaller than the values measurable
or invisible of the neck size. In practical manufacturing, only visible defects are monitored. For
this reason, the formulation of plastic instability criteria FLC always underestimates the tube’s
visible formability and is over conservative to evaluate the forming severity of parts.
Figure 5.4 Effects of the t value on forming limit curve with Keeler’s formula.
a.) Results of the formed product for experiment and numerical simulation
b.) Comparison of major and minor strain for experiment and numerical simulation
Figure 5.6 Results of comparison for experiment and numerical simulation
During simulations, a elastic-plastic material model considering strain hardening
obtained from tensile tests is used. Symbols (X) and (O) represent the major and minor principal
strains of the mesh where the real experiment and numerical simulation (Finite Element software:
DYNAFORM), respectively. It can be seen that simulated strains closely follow the measured
strains. Therefore, FE model can reasonably simulate hydroforming process.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
-0.3 -0.25 -0.2 -0.15 -0.1 -0.05 0 0.05 0.1
Experimental FLC
Experiment
Simulation
Burst point
Major
strai
n (ε 1
)
Minor strain (ε2)
64
5.2.2. Verification of Experimental FLC with a real automotive part
The forming limit curves determined for these tubular materials were put to test
in formability evaluations of test parts forming in both real experiment and numerical simulation
(Finite Element software: DYNAFORM ). A real automotive part, i.e. a fuel filler pipe (see
Figure 5.7), was considered in this study to design proper process parameters.
First, FLC of comparable steel sheet available in DYNAFORM was used to
conduct all the simulation runs and burst, shown in Figure 5.8.
Figure 5.7 A fuel filler pipe geometry.
Figure 5.8 A final product of fuel filler pipe.
An actual fuel filler pipe load path was put to test in numerical simulation (Finite
Element software: DYNAFORM). The results were found that the final tube in numerical
simulation is closely shaped with the formed tube in real experiment as show in figure 5.9. Figure
5.10 provides a correlation between data points measured experimentally from strain grid of fuel
filler pipe. The predicted FLCs provided from plastic instability criteria and Keeler’s formula is
plotted in the true axial strain versus the true circumferential strain space. The experiment data
65
and simulation data of strain grid analysis for fuel filler pipe are plotted as discrete points in the
same figure. As observed in Figure 5.10, the experiment data is close to simulation data and close
to the experimental FLC. It can be seen that all the measured strains near the crack site are located
above the experimental FLC. Also, none of the critical strains (Crack) lie between the
experimental FLC and Keeler’s FLC. Therefore, from this comparison, the experimental FLC
seems to be the best FLC to estimate a necking (i.e. crack) of this particular tubular steel material.
Figure 5.9 A simulation model of fuel filler pipe.
Figure 5.10 Comparison of predicted forming limit strains with measured experimental data.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
-0.3 -0.2 -0.1 0 0.1 0.2 0.3
Keeler’s formula
Formulation of plastic instability criteria
Experimental FLC
Automotive part(Real Experiment-cracks)
Automotive part(Real Experiment-safe)
Automotive part(Numerical simulation )
Major
strai
n (ε 1
)
Minor strain (ε2)
CHAPTER 6
CONCLUSIONS AND SUGGESTIONS
6.1 Conclusions
In this study, FEA was used to investigate and determine proper FLC testing die insert
geometry, i.e. die entry radius and bulge length, in relation to tube sample geometry, i.e. outer
diameter and tube thickness. It was found that the proper die geometry does not depend on sample
thickness but only on outside diameter of the sample. A FLC testing apparatus has been designed
for commonly used STKM 11A tubing in Thailand. Testing process parameter (loading profiles)
were also determined for conducting the test that guarantees linear strain paths in the deforming
samples. This knowledge can be applied to design proper FLC testing die inserts for other tubing
dimensions. Forming limit curves to be generated using this testing apparatus will be of great
usefulness for Thai industry in designing and producing high strength-to-weight ratioed parts
using tube hydroforming technology.
FEA was also used to determine the loading paths to be implemented by the hydraulic
machine with axial feeding and pressure followed the prescribed loading paths that correspond to
the strain paths with a constant strain ratio at the pole of the forming tube. A forming limit
diagram from forming limit experiments was successfully using an experimental apparatus with
fixed bulge length and a test machine with axial feeding. Analytical forming limit curves were
also constructed using Swift’s diffused and Hill’s localized necking criteria associated with Hill’s
non-quadratic yield function, and Keeler’s formula. From the comparison between the
experimental, analytical, empirical FLCs, it was concluded from using a real THF part that the
experimental FLC seemed to be able to predict the necking best. The main conclusions from this
work can be made as follow.
- A bulge forming apparatus of fixed bulge length and a hydraulic test machine with axial
feeding are designed and used to carry out the bulge tests.
- The loading paths of four different strain paths can generate strain paths with a linear
strain path at the pole of the forming tube.
- The forming limit curve (FLC) of tubular material low carbon steels commonly used in
Thai industry, namely STKM 11A is determined
67
- The forming limit curve (FLC) is verified with real part forming experiments for actual
application accurately seems to be not appropriate for THF.
- The welding line quality of ERW tube in Thai industry is low quality.
6.2 Suggestions for Future Work
Although this work is finished, there are some suggestions to the work for further study
determination of forming limit curves of tubular materials for hydroformability evaluation of
automotive parts
- Study effect of weld line expandability on formability in hydroforming of ERW tubes.
- Improvement of steel tubing production toward proper usage for THF technology.
REFERENCES
1. Ahmetoglu M. Tube hydroforming: current research, applications and need for training.
Journal of Materials Processing Technology 98 (2000) 224-231
2. Altan T. Evaluation of tube formability and material characteristics. Journal of Materials
Processing Technology 98 (2000).
3. Asnafi N. Theoretical and experimental analysis of stroke-controlled tube hydroforming.
Materials Science and Engineering A279 (2000) 95–110
4. Chen J. Sheet metal forming limit prediction based on plastic deformation energy. Journal
of Materials Processing Technology (2009)
5. Chu E. Influences of generalized loading parameters on FLD predictions for aluminum
tube hydroforming. Journal of materials processing technology 196 (2008) 1–9
6. F. Dohmann Hydroforming - a method to manufacture light-weight parts. Journal of
Materials Processing Technology 60 (1996) 669-676
7. Hwang Y.M. Analysis of tube bulge forming in an open die considering anisotropic effects
of the tubular material. International Journal of Machine Tools & Manufacture 46 (2006)
1921–1928
8. Hwang Y.M. Forming limit diagrams of tubular materials by bulge tests. Journal of
Materials Processing Technology (2009).
9. Kang S.B. Analytical and numerical approach to prediction of forming limit in tube
hydroforming. International Journal of Mechanical Sciences 47 (2005) 1023–1037
10. Koc M. Hydroforming for advanced manufacturing. Woodhead publishing in materials
11. Koc M. On the characteristics of tubular materials for hydroforming experimentation and
analysis. International Journal of Machine Tools & Manufacture 41 (2001) 761–772
12. Koc M. Prediction of forming limits and parameters in the tube hydroforming process.
International Journal of Machine Tools & Manufacture 42 (2002) 123–138
13. Lei P.L. Bursting failure prediction in tube hydroforming processes by using rigid–plastic
FEM combined with ductile fracture criterion. International Journal of Mechanical Sciences
44 (2002) 1411–1428
14. Nefussi G. Coupled bucklingand plastic instability for tube hydroforming. International
Journal of Mechanical Sciences 44 (2002) 899–914
69
15. Sonobe O. Effect of Mechanical Properties on Formability in Hydroforming of ERW
tubes. JFE STEEL Corp., Steel Research Labs., JAPAN
16. Takuda H. Finite element analysis of limit strains in biaxial stretching of sheet metals
allowing for ductile fracture. International Journal of Mechanical Sciences 42 (2000) 785-798
17. Takuda H. Prediction of forming limit in bore-expanding of sheet metals using ductile
fracture criterion. Journal of Materials Processing Technology 92-93 (1999) 433-438
18. Takuda H. The application of some criteria for ductile fracture to the prediction of the
forming limit of sheet metals. Journal of Materials Processing Technology 95 (1999) 116-121
19. Vollertsen F. On possibilities for the determination of the coefficient of friction in
hydroforming of tube. Journal of Materials Processing Technology 125-126 (2002) 412-420.
20. Xing H.L. Numerical analysis and design for tubular hydroforming. International Journal of
Mechanical Sciences 43 (2001) 1009}1026
21. Yannis P. Inflation and burst of anisotropic aluminum tubes for hydroforming
applications. International Journal of Plasticity 24 (2008) 509–543
22. Yoshida K. Effect of strain hardening behavior on forming limit stresses of steel tube
subjected to nonproportional loading paths. International Journal of Plasticity 23 (2007)
1260–1284
APPENDIX A
2nd INTERNATIONAL CONFERENCE ON GREEN
AND SUSTAINABLE INNOVATION 2009 Appendix A-1: Determination of forming limit curves of tubular materials for
hydroformability evaluation of automotive parts
71
72
73
74
75
76
77
78
APPENDIX B DATA ANALYSIS
Curve Correction
𝑙 = 2𝑟 sin
𝑙 = 2𝑟 sin
Where 𝑙 = Major deformation 𝑙 = Minor deformation 𝑙 ′ = Captured major deformation 𝑙 ′ = Captured minor deformation 𝑟 = Meridian radius of curvature at the pole 𝑟 = Circumferential radius of curvature at the pole
Strain
ε = ln
ε = ln Where ε = Major strain ε = Minor strain d = Initial grid length